Master Thesis: Comparison of Bayesian Estimators for random access UWB positioning

What to expect

Ultra wideband (UWB) positioning enables accurate distance estimation between wireless nodes by measuring signal propagation times. A widely used approach is Two Way Ranging (TWR), in which the exchange of multiple messages between two nodes allows them to estimate their distance with high precision while compensating for clock offsets and drifts. Conventional TWR between more than two nodes is often realized using time division multiple access (TDMA). However, TDMA comes with challenges in scheduling messages in short sequence. Applications which require high measurement frequency make random access (RA) approaches particularly attractive, as they avoid these scheduling challenges and also offer greater flexibility in dynamic networks.
The main goal of this thesis is to compare RA based Bayesian positioning approaches. The first approach follows the classical two Step approach, in which range measurements are computed from TWR message packets and then used as inputs for a positioning filter. The second approach works directly on the recorded timestamps of the exchanged ranging messages and jointly estimates the nodes’ positions and relevant clock parameters.

Your tasks

• Compare distance-based and time-based Bayesian estimation approaches for different time scales and system dynamics
• Analyze UWB random access packet loss, channel load, and their impact on estimation performance
• Work with both simulations and real-world data
  

Your profile

• Enrolled in Electrical Engineering, Computer Science, Physics, or a related field
• Strong background in estimation theory, signal processing, and applied mathematics
• Experience in software programming, e.g. Python, C++, MATLAB, or similar
• Independent and structured working style
• Good written and spoken English or German